From idealized clinical empathy to empathic communication in medical care

Part one: conceptualizing clinical empathy

A brief history of “detached concern”

Outside of healthcare settings, people use the term empathy to mean “feeling with” another person or putting yourself in someone else’s shoes. The assumption is that emotional resonance with another clues you in to how they feel. This has been problematic for doctors, who have historically believed that they could understand their patients’ feelings while striving for emotional detachment: objectivity is seen as crucial for making tough diagnoses, and stoicism is believed to be necessary for providing invasive, sometimes noxious, treatment. Additionally, there is concern with avoiding burnout, or more specifically, compassion fatigue.¹

Thus doctors have aimed for their own idealized version of empathy, one in which they suppress personal emotions yet are motivated by an altruistic yet “detached” concern for patients. The term “detachment” can mean different things, some of which do not involve suppressing emotions but rather accepting them. For example, for Zen Buddhists “detachment” refers to allowing emotions to come and go without attaching to any particular emotion. In contrast, American physicians from the early twentieth century until the late 1960’s used it to mean suppressing emotional responses.² Writing in 1906, Sir William Osler, father of “modern medicine,” describes how only if the physician was emotionally “imperturbable” so that his “blood vessels don’t constrict and his heart rate remains steady when he sees terrible sights” will he have the “equanimity” to “see into” the patient’s “inner life.”³

Once emotions are suppressed, what is the basis of the concern for patients? The “concern” in “detached concern” was not based on untrustworthy feelings, but rather on a duty or commitment to heal. Note that Osler acknowledges that physicians are prone to sympathetic identification with their patients, but hopes that they will avoid feeling sympathy and be moved by a purer professional attitude:

The more closely we [the physicians] study their [the patients’] little foibles of one sort or another in the inner life which we see, the more surely is the conviction borne in upon us of the likeness of their weakness to our own. This similarity would be intolerable if a happy egotism did not often render us forgetful of it. Hence the need of an infinite patience and of an ever-tender charity toward these fellow-creatures.⁴

While Osler’s views were influential in the early twentieth century, it was in the 1950s and 1960s that “detached concern” became the overarching ideal for medical professionalism.⁵ Physicians, thankful for the scientific progress that was finally making medical care effective, believed that norms like objectivity should extend even to their behavior towards patients. Accordingly, “detached concern” was an idealized, white-coated concern, in which by refraining from emotional contamination, physicians would gain therapeutic power. Observing medical students, sociologists Renee Fox and Harold Lief write:

The same detachment that enables medical students to dissect a cadaver without fear or disgust seemingly enables them to listen to patients empathically without becoming emotionally involved.⁶

By describing this extreme reversal of the ordinary concept of empathy, Fox and Lief were also calling “detached concern” into question. Physicians during this era recognized that they had a range of emotional responses to their patients, including empathy, sympathy and other caring attitudes, yet when writing major articles for leading medical journals they emphasized the risks of any emotional engagement and the benefits of detachment.⁷

Unfortunately, patients did not see detachment as beneficial. As physicians in the 1960’s strived for detachment, patients, nurses, clergy and others began to question doctors’ attitudes towards patients. Alongside other populist anti-authority movements of the late 1960’s and 1970’s, the American bioethics movement arose in protest against both medical paternalism and failures of compassion in medical care.⁸ There were dramatic media examples of dehumanizing medical researchers whose behavior other doctors seemed to accept. For example, Chester Southam, the same physician who was placed on probation for injecting live cancer cells into elderly patients at the Jewish Chronic Disease Hospital in Brooklyn, New York, was elected Vice President of the American Cancer Society 2 years later.⁹

Cases like this led patients to question the adequacy of detached professionalism and to demand more recognizable empathy from their physicians. Yet until the 1990’s there was almost no actual research on what role empathy played in effective medical care. Since then, and especially in the past few decades, medical educators have taken a more complex view of empathy.

Rethinking empathy in light of research

This section delineates a model of clinical empathy in light of conceptual and empirical research. In the next section we consider how implementing this model is not always realistic or even best in every medical context, suggesting the need to study empathies in context rather than to defend a singular ideal for clinical empathy. This contextualized approach also invites a shift from focusing on the internal mental processes of the empathizer to the dynamic interaction between two people engaged in empathic communication.

Conceptualizing empathy

While the term “empathy” has many distinct meanings (Daniel Batson gives at least eight)¹⁰ –the term “clinical empathy” has been conceptualized more specifically. The author, integrating the empirical research of many others, has argued that the medical setting calls for both cognitive curiosity/perspective-taking and affectively attuned empathic communication, and that the two interact and support each other.¹¹ Building from this work, and more recent research, let us begin with the assumption that the goals of clinical empathy are two-fold: to understand the patient’s experience sufficiently to be effective in treating his or her illness; and to communicate successfully so as to build a good therapeutic alliance, necessary for effective treatment. These two goals can be distinguished as cognitive and communicative, for shorthand, although they support each other in practice.

Conceptually, empathic understanding and empathic communication are distinct goals, and practically, there are many contexts in which people engage in the former without the latter—consider an advertising agency studying their target audience’s desires in order to sell a product. Conversely, empathic communication at some level is possible without understanding. Person A can resonate with person B’s sad feelings in a way that B can read, without A ever understanding what specifically B is sad about. The fact that cognitive empathy, and affective communication can function independently is supported by neuroscience as well, a topic we turn to shortly.

However, observational studies (see next section) show that in medical practice, the cognitive and affective aspects of empathy support each other. The definition of clinical empathy that I will thus argue for is engaged curiosity, in which the clinician’s cognitive aim of understanding the patient’s individual perspective is supported by affectively engaged communication.

Observational research in medical settings

Basic psychology and neuroscience research have, due to methodological and other factors, restrictively interpreted the “cognitive” aim of empathy as corrrectly labeling another person’s emotion type.¹² Unfortunately, this interpretation has been adopted by the discourse on medical empathy, which too often focuses only on the clinician’s correctly identifying the patient’s emotion. However, it is often quite obvious that a patient is sad versus angry, and the crucial aim is to learn what in particular is bothering this individual. Often this involves imagining his or her emotional point of view in some detail: what is it about accepting a new diagnosis or trying to change a habit that most depresses or worries this person? Importantly, clinical research suggests that to meet this more complex cognitive aim affective resonance or non-verbal communication plays an important role.

For example, observational studies show that patients give more information to emotionally attuned physicians than to those who ask good questions but are not emotionally engaged. Researchers in the United States, Norway and elsewhere have performed observational studies of patient-physician interactions that show that before patients talk about aspects of their history that are emotional, they give hints, often through gestures.¹³ When physicians respond to these cues with questions but little non-verbal attunement, no disclosure takes place. In contrast, when physicians show non-verbal attunement at these critical moments patients give fuller histories. Better-informed physicians are more likely to correctly diagnose and effectively treat medical problems.

In addition to improving diagnoses, empathic communication makes medical care more effective. Perhaps the biggest barrier to effective treatment is patient non-adherence; in some studies about fifty percent of treatments are not taken as prescribed. Some time ago, Debra Rotor of Johns Hopkins performed a meta-analysis of factors that improve adherence.¹⁴ She found that trust was a most important predictor of adherence. She then asked what factors predict trust. Physician friendliness did not engender trust, but the patient’s sense that the physician was genuinely worried about him or her did predict trust. Since Rotor’s research, other studies have shown that patients trust empathic physicians more, and that this increases the effectiveness of medical care.¹⁵

Empathic communication plays other roles in effective treatment. Studies show that it helps empower patients to address their medical problems. Research on giving bad news shows an important role for emotionally engaged listening in helping patients cope when hearing a serious cancer diagnosis. Emotionally engaged physicians communicate better with patients, decrease patient anxiety and increase patient coping skills, leading to better outcomes. In particular, patients were better able to actively engage in and make cancer treatment decisions and to seek out support when their physicians were more emotionally engaged. ¹⁶

Cutting edge research connecting empathy and medical effectiveness suggests that in addition to empowering patients to care for themselves, empathy may play a direct role in healing. Studies not only link the patient’s perception of empathic care with their taking better care of themselves, including better control of blood sugar and cholesterol,¹⁷ but also suggest links between empathy and increased immune responses to and reductions in duration of the common cold.¹⁸ Physicians have been calling for research in psychoneuroimmunology to elucidate the mechanisms of the mind–body healing they observe in their practices.¹⁹

What does neuroscience teach us?

In addition to these observational studies in medical settings, recent neuroscience research addresses some questions about empathy, while raising others. Research in psychology has tended to presume a competing view of the affective and cognitive aspects of empathy, pitting one against the other.²⁰ Neuroscience seems to support this view insofar as it has demonstrated that different parts of the brain light up for affective versus cognitive cues, thus establishing distinct pathways.²¹ Note however that this does not yet tell us how these pathways do or do not work together in naturalistic contexts.

This competing view has colored the interpretation of an important finding regarding clinical empathy. Jean Decety and others found that physicians’ brains show an abnormal lack of response to physical pain cues; while a non-physician seeing a needle going into skin responds with a similar brain pain pattern as if he were feeling the needle himself, physicians show much less of this response. Physicians simultaneously showed more brain activity in areas involving executive control and self-regulation.²² These findings are not so surprising and make sense insofar as physicians must become comfortable performing painful procedures. One implication of such findings is that it is important for physicians to use cognitive empathy when communicating with patients about painful procedures to ensure that the patient’s needs are being adequately addressed.

However, the fact that physicians have reduced empathy for physical pain does not entail that they have tamped down responses to emotional pain. Yet it does seem important for physicians not to become too anxious in response to their patient’s suffering. Research shows that medical students who become more personally distressed in response to patients’ distress had steeper declines in cognitive empathy during training.²³ Combining findings about personal distress with the pain studies has led some to wonder if students with more emotional empathy will lack cognitive empathy. However, this overlooks important distinctions between self-related anxiety and affective resonance with another.²⁴ While resonance can be a trigger for self-related distress in some physicians, others can respond with emotional empathy without becoming personally distressed. Notably, physicians who are skillful at emotionally engaged empathy—who can sustain affective attunement without becoming too anxious—appear to have the highest level of well-being and job satisfaction, even when practicing in such fields as palliative care which offers frequent triggers for less interpersonally skilled physicians.²⁵

Further, as noted in the observational studies above, many physicians seem able to attune non-verbally to patient’s anxiety, fear and grief, suggesting affective as well as cognitive empathy in such cases. Presumably, having a tamped down response to seeing a needle going into skin does not predict being less responsive to a patient’s grief. We need more research on the neural pathways involved in more complex emotional communication. Understanding another person’s grief requires some cognitive empathy insofar as seeing their embodied affect as grief involves understanding what their emotion is about—for example, the loss of a loved one. Yet it is likely that affective resonance also plays a role.

Recently, outside of the medical setting, neuroscientists Jamil Zaki and others have argued that it is crucial to recognize that existing brain studies are subject to certain biases towards segregating and oversimplifying affective and cognitive empathy, whereas it is likely that in naturalistic settings they interact in complex ways. This is in part because the targets involved in such research are usually the kind of stimuli that call for segregated responses–like facial expressions versus written word narratives. In contrast, real life interactions involve multimodal cues.

Following Zaki, we can hypothesize that for physicians, as for non-physicians, which pathway is elicited depends on the context. For example, sometimes the information a patient gives is more narrative in form, other times more affective and non-verbal. Just because physicians automatically tamp down responses to needles in skin does not entail that they automatically tamp down responses to non-verbal as well as verbal expressions of grief. This needs to be studied.

Part two: implications for medical practice and education: from empathy to empathies

Bringing together theoretical, observational and neuroscience research leads to the hypothesis that both cognitive and emotional empathy play important roles in clinical contexts. We need further research to elucidate when and how the two processes work best together versus apart. Perhaps even more importantly, it is time to shift our focus from describing the fullest model of clinical empathy under ideal circumstances to studying the range of empathic processes or “empathies”—cognitive, affective and communicative subcomponents of empathy—that are practical in different clinical contexts. Finally, while experiencing full-blown affective-cognitive empathy is not under our direct control, clinicians can consciously cultivate empathic practices. Recent research findings that might help doctors to sustain such practices are discussed in the final section of this paper.

From empathy to empathies

So far we’ve suggested that clinical empathy involves cognitive and affective processes working together to guide not only understanding but communication. We’ve defined full-fledged clinical empathy as affectively informed engaged curiosity. Other work beyond the scope of this paper describes how it is possible for emotions to inform empathic understanding by guiding imaginative processes.²⁶ The focus here, however, is not on how empathic understanding itself operates in the mind of the empathizer, but on how bilateral empathic communication operates in medical care. Once we shift our attention from the mental faculties involved in empathic understanding to empathic communication, we can attend to particular contexts. We can shift from asking what is clinical empathy to asking what are the variety of empathies that improve clinical care?

How practical is it to expect empathic communication given the demands of medical practice today—excessive work load, tremendous time pressure, unfamiliar patients, and duties that vary from comforting to performing invasive procedures? What is practical, and often appropriate, is limited or partial empathic engagement. Sometimes empathic communication is primarily cognitive for both parties, sometimes more emotional for both, and often more cognitive for physicians and more emotional for patients. In fact, in educating young clinicians, I urge them not to define one thing as “clinical empathy,” but rather to notice the variety of ways that they empathize and that their patients respond. We should develop a research agenda to study “empathies” in different contexts depending on: what is the clinical focus (performing surgery, giving bad news, encouraging weight loss); what cultural/historical factors including previous medical experiences shape the expectations of both patients and physicians; what is the evolving tone of this specific clinical interaction?

Such contextual factors determine whether or not common expressions of empathy, or their absence, genuinely serve or interfere with good communication. In many contexts it is simply unnecessary for clinicians to engage in any kind of empathy beyond acknowledging the patient as a human being with feelings and worth. A patient seeing a primary care physician for a simple health problem like a strep throat may need a remedy and nothing more. In some contexts patients may want their physician to understand their perspective cognitively but not to be too engaged emotionally—for example, a reserved teenage boy giving a sexual or substance use history might perceive an emotionally neutral physician as non-judgmental and thus more trustworthy.

While context is crucial, the practice of empathy cannot be based on applied sociology. We cannot codify and prejudge typical clinical settings or types of social interactions in terms of empathy demands. A large percentage of primary care visits that are first called in as mundane physical problems are motivated simultaneously by the patient’s underlying depression, anxiety, alcohol or drug concerns or domestic violence.²⁷

Studies show that in the context of giving bad news or discussing other serious losses, conveying some kind of appropriate emotional response—studies label it “empathy,” but it could be attunement or concern—serves patients.²⁸ However, this is not always the case. I have seen expressions of concern make some patients very uncomfortable. For example, I have previously described a case in which I first met with a powerful businessman in his late fifties who was in the ICU because a rare disorder (Guillan-Barre) left him suddenly paralyzed from the neck down. Seeing him splayed in his hospital bed, helpless and exposed before his family, led me to approach him gently and, I later realized, sorrowfully. Unfortunately, my obvious sympathy for his condition made him feel more shame, and he lost interest in talking with me. When I returned to meet with him later on, I deliberately took on the firm, assertive tone of a business interaction and asked him to tell me what he didn’t like about talking with me. This allowed him to yell at me, directly expressing his anger at not only his medical doctors (including me), but his tragic situation. When he saw that I could calmly listen and show interest in his anger he began to trust me and was then ready to share his more vulnerable feelings and work with me.

Note that some might say that no actual empathic resonance or cognitive role- taking was required. I could have typified the patient as someone who would prefer a business-like approach form the start, and perhaps speaking to him firmly would be enough to put him at ease. This applied sociology approach is supported by those who argue that physicians primarily need to learn how to “act” (as in acting on the stage) in ways that patients will perceive as empathic.²⁹ However, how was I then supposed to act when the patient yelled at me? In this case, calmly accepting his anger was correct, but in others it is not. In my view, he developed trust in me not so much because I got it right so much as because I kept trying, and he could see that I was listening to him and truly attentive to his emotions as they shifted in real time. This is not something that can be performed. So on the one hand, I did “act” when I consciously shifted my manner towards that of a business interaction, on the other hand, I was non-verbally attuned to him when he expressed his anger and ultimately his grief.

Thus in arguing for the practicality of “partial” empathies, I do not endorse the current trend to encourage superficial acting or other routinized approaches to substitute for genuine empathy. However, we need research to elucidate what conscious or “acting” behaviors can help induce genuine empathic communication between patients and physicians.

Why engaged curiosity is necessary

Physicians talk to so many patients that they will inevitably make communication mistakes as I did initially in the case above, yet they can still develop a therapeutic alliance by sustaining their interest in the patient’s perspective. I refer to this interest as engaged curiosity. The basic stance is one in which the physician recognizes that he or she does not fully understand and has more to learn about the patient’s situated experience—what the patient is afraid of, needing or seeking. While this may sound simple, it is not, for it presumes that the physician maintains awareness of the distinction between herself and the patient, avoiding errors of merging (“We are in the same boat”) or projection. To help medical students remember this, I urge them to ask the patient to “tell me what I’m missing” and to try not to say “I know how you feel.”

Engaged curiosity is especially important in emotionally distressing situations. Whether or not physicians try to empathize with their patients, they are in fact often deeply affected by the suffering and emotional difficulties they witness. In addition, as is discussed later in this article, doctors can get into conflicts with patients and families, and become frustrated and angry. Physicians who are unaware of their own emotional state risk making poor decisions to alleviate their own distress. Therefore becoming interested in and reflective about the implicit emotional communication between patients and themselves can be important for sustaining professional conduct.

The following cases show how serious an impact unrecognized emotional responses to patients, including unreflective sympathy, can have on medical care. An article from Seminars in Pediatric Surgery describes a problematic lack of curiosity about the patient’s feelings, not only from the clinicians involved but even from the ethics experts who wrote the article.³⁰ A baby is born at thirty-two weeks with severe neurological impairments, among other problems; doctors “prematurely” discuss not only the infant’s death, but organ donation and burial. Returning to the hospital a year later when the child is gravely ill, the parents seek all possible care. They are angry and distrustful towards the pediatric surgeons who believe that treatment should be terminated. The mother angrily expresses her distrust towards the physicians, saying that, by living for a year, “my son proved you wrong.” She refuses to comply with their treatment recommendations. The authors acknowledge that the problem here is distrust over the previous communication mistake, but they see no recourse other than to transfer the child to another hospital.

What I find most interesting is what the authors do not suggest: the surgeons might apologize for their earlier poor communication with the parents. Presumably, the mother’s angry comment that her son proved the doctors wrong might have been met with an apology for barraging her with bad news prematurely. This acknowledgment would have capitalized on the opportunity for a psychological connection that the mother created by expressing her anger directly–she was actually courageously reaching out. Most likely, her expressed anger at the physicians was a defense against her much more intolerable feelings of anger about having a severely ill child. It seems likely that this mother needed emotional support rather than a new team of doctors at another hospital.

I was not directly involved in the case above and cannot speak to the states of mind of the physicians, but I suspect that they were not simply callous but rather anxious about this case, and that their insensitivity resulted from failure to adequately think about the mother’s perspective. That is, I hypothesize that their failure was primarily one of cognitive empathy. They might have experienced some resonance but they did not express curiosity about the specific needs they could address for this mother. Rather, they acted unconsciously to relieve their own distress.

In a very disturbing case that I was involved in, which I have been examining and developing new ideas about for more than a decade, I can attest to just such a process. This is the case of Ms. G, a woman in her fifties who comes out of surgery for a second leg amputation (related to the complications of diabetes) to find that her husband of over twenty years is suddenly leaving her. His words are particularly selfish and cruel—he tells her that with her amputations and other health problems he could not be attracted to her–and he has already started a new relationship with someone. Ms. G responds with suppressed rage, shame and an overwhelming catastrophic view of her future as one in which she will forever be alone, a pariah, forever feeling her present sense of abandonment. She feels overwhelming pain, an admixture of physical and mental anguish, and refuses life-sustaining dialysis knowing that she will die.

While it is completely understandable that Ms. G might feel like dying at such a moment, the problem in the case is that her entire team of physicians quickly agrees with her and within one day, they turn up the morphine so that she lapses into deep sleep and dies from (predictable) respiratory arrest. How did this happen?

Her caregivers were subject to mostly unconscious, self-related anxiety, or “sympathetic distress” which impelled them to act urgently to decrease their distress. Humans are empathically resonant animals; we are almost always affected by another person’s distress. While doctors may become desensitized to putting needles into skin, they are not necessarily desensitized to seeing people suffer from human cruelty, even if they work in prisons where they are exposed to terrible stories on a daily basis.³¹

Patients’ suffering does not merely take place somewhere in their “heads” but rather shows up in the emotional communication between members of the treatment team. People who are suffering tend to convey their suffering through their interactions with others. As adults we are often unaware of how we reenact emotionally upsetting events. But in the case of children the reenacting of painful experiences is easy to observe. This is in fact the basis of therapy to help children recover from terrible losses. When children are simply invited to play, they act out their most painful experiences. A therapist empathizes with the child by recognizing what is being expressed in the play, in such a way that the child can directly feel fear or grief in the presence of the therapist, often for the first time.

Adults repeat their traumas just as children do, and in the intense setting of the hospital, it is not unusual for the whole treatment team to engage in the reenactment without being aware of it. Ms. G’s medical team unthinkingly adopted her feelings of hopelessness and helplessness and thus urgently hastened her death, reenacting her abandonment trauma.

What should happen instead? Physicians need to cultivate curiosity about their own emotional reactions and about what they might be missing about the patient’s experience that, if better understood, might help them address the patient’s suffering. First, while all of Ms.G’s caregivers were clearly deeply affected by her we never discussed how emotionally affected we were, and we never reflected about whether our hopeless feelings were grounded in facts or just our own resonance with her distress. The failure to reflect was notable, since a week before the doctors’ notes in her charts all described her as a good candidate for surgery, someone who could expect a reasonable quality of life in the future. Yet we never questioned our sudden 180 degree turn to feeling completely hopeless about her future—two of her physicians said that her future was truly hopeless, that they would want to die if they were her. Given this unreflective hopeless feeling, we ignored several pieces of clinical evidence that predicted that she could recover emotionally and adjust to a new life. She had previously felt just as hopeless about her future when her first leg had been amputated, but she had recovered from what in retrospect was a transient period of depression. She had a very strong underlying character and had shown great resilience in dealing with losses over her lifetime. She was at the height of her career as an artist, and her medical problems did not prevent her from pursuing her art. She had very supportive friends who were holding a daily vigil outside her door, even though she was temporarily unwilling to have them speak with her.

It was only right then, in the immediate aftermath of her husband’s cruel rejection that she felt catastrophic about her future, yet we joined in with her catastrophic feeling unthinkingly. We would never have simply adopted a patient’s biologic self-diagnosis without thinking analytically about her medical history, but when it came to emotional predictions about the future we lacked reflectiveness and accepted the patient’s views at face value.

We not only failed to think about her conviction that her future was now hopeless, we failed to recognize how anxious we were all becoming, and were unaware that this anxiety was driving us to urgent action, including ultimately ending Ms. G’s life prematurely. When she rejected all of our attempts at connecting we felt helpless as well as hopeless and abandoned her. For example, I left her room when she yelled at me because I felt too helpless to calm her down.

This case illustrates the limitations of affective empathy in the absence of cognitive empathy/curiosity and perspective-taking. Affective resonance is an unthinking response. Resonating with another’s expressed emotions is one thing. It is quite another to keep in mind that the most intensely expressed feeling may not be a person’s only or even most important emotional attitude. Human emotions are complex, we are often subject to emotional ambivalence. Thus when a patient sees the world in a totalizing way, as in catastrophic thinking, or, for that matter, wishful thinking, our job is to become curious and interested about what lies behind his or her apparent certainty. We might ask: “You feel that your future is over, but what makes you so sure?” or “You’re sure you won’t get diabetes from being overweight, but let’s consider if you also have some concerns.” For someone like Ms. G, it seems clear in retrospect that she was frightened and angry, yet we failed to explore this and simply adopted her apparent hopelessness as her singular possibility.

Ms. G was in a moment of crisis that made it appear that her feelings had no complexity, but our emotional lives are always complex and open to new discovery. We might have been curious about how her whole life perspective was collapsed at the present moment, and helped her become aware that this was the case. For example, given how important her artwork and friends had been to her, there was no evidence that her marriage was of such overriding value that she would have chosen to die if, for example, her husband had died. Perhaps Ms.G was subject to overwhelming anger and shame at being so cruelly rejected, and she wanted to die immediately to show her husband and everyone else that she would not tolerate such rejection, and if this were the case these feelings might have masked other desires to go on with her life.

Cases like the above show that physicians need to be more reflective about how their patients’ emotions are influencing their own clinical judgment, even if they are not aware of being emotionally affected. When we feel an urgent need to act we should become curious about what is making us anxious. Accepting that unconscious emotional reactions influence our apparently professional judgments and decisions also helps us avoid projecting our own personal issues and concerns onto patients, and helps us sustain appropriate clinical boundaries.

To help physicians learn about this, they need to be educated about how emotions are not just embodied feeling states but implicit judgments about situations—for example, to become angry at someone is already to believe that that person is doing something harmful or wrong. Emotional responses to patients like Ms. G influence not only how members of the treatment team feel, but what they think. Emotional responses involve implicit judgments, and these judgments are not always adequately based on facts or evidence.³² The team treating Ms. G felt so helpless witnessing her predicament that we saw her situation as truly dismal, and lost sight of her very realistic chance of significant recovery. We might instead have become curious about how her sense of the future was collapsed at the present moment, and helped her become aware that this was the case.

Thus my recommendation is that the kind of empathy that patients most need from us is empathic curiosity—genuine, emotionally engaged interest in learning more about the complexity of their (and our responsive) emotional points of view. Engaged curiosity depends on brain pathways for cognitive empathy, but also engages affective pathways. The goal is to utilize affective resonance or feelings of concern reflectively, in the service of better understanding and communicating with the patient.

Dispelling myths and misunderstandings to help physicians sustain empathy

I hope that the discussion thus far has shown the clinical importance of empathic curiosity and motivates us to value it. However, you may now be asking: how can all this be practical? Given how busy clinicians are, how can they cultivate such curiosity, especially in difficult situations? And how can they avoid feeling badly about themselves when they don’t feel empathic interest in the patient? Here are some surprising findings that may alleviate some of these concerns.

First, finding it difficult to empathize during a conflict about treatment does not mean that you are an uncaring doctor

Actually, it can mean just the opposite. When patients like Ms. G refuse medically necessary care, physicians are left feeling responsible for the patient and yet impotent to help him or her. This is precisely the kind of self-related emotional distress that makes it difficult to focus on the other person’s perspective. Physicians who feel most responsible for their patients are more likely to find such situations frustrating.³³ Further, basic psychological research suggests that people who are more concerned about others might be particularly bad at focusing on other’s perspectives during conflicts.³⁴

There may be cultural or other interpersonal differences that make it hard for an individual physician to read a particular patient. Note that in trying to imagine patients’ experiences, it is important to recognize that we often cannot do so fully or accurately. Individuals vary in their ability to imagine another person’s thoughts and feelings, and some people may be easier to “read” than others. However, the good news is that it appears that simply making the effort to understand the other person’s perspective plays a helpful role in conflict resolution.

Second, paying more attention to our own feelings of discomfort and anxiety will lessen, not exacerbate those feelings

In Ms.G’s case the whole team failed to take our emotional temperature. But if we had, we could have then taken steps to lessen our anxiety, rather than accepting Ms.G’s request to stop treatment so quickly; what was needed was the patience to tolerate emotional distress.

Recently, some medical schools have begun to teach physicians rapid meditation techniques that can help them reduce their own anxiety. Painful emotions ebb and flow, and seeing this ebb and flow can lessen the sting. These interventions have helped doctors stay emotionally engaged with patients while feeling less distressed. Even a few minutes of self-awareness has been shown to reduce errors, improve decision making, and resolve conflict.³⁵

Third, while the stereotype is that people avoid negative feelings, we are built to be much more interested in our negative than our positive feelings

Research suggests that negative emotions besides rage (and extreme states of fear or despair) tend to make people more inquisitive about the basis for their emotional views, seeking more information than people do when they are in positive emotional states.³⁶

Of course it is one thing for people to be curious about their own emotions and another for them to become curious about another person’s, especially when that person is a component of the distress. However, physicians can learn to bridge the two by becoming interested in how their own negative emotions provide important clinical clues to what the patient might be feeling. Using one’s own emotions as clinical information is an important part of psychotherapy training. For example, when a therapist feels ongoing hopelessness or frustration towards a patient, she considers possibilities including whether the patient might feel similar or complementary feelings.

A fourth surprising observation is that the less a clinician tries to say something smart or knowledgeable about the patient’s psyche, and the more he or she simply repeats the patient’s exact words, the more the patient will communicate

Skillful empathic listening doesn’t depend on psychotherapeutic techniques like knowing how to make subtle interpretations of patient’s inner lives. Rather, listening closely enough to someone to repeat his or her exact words is an effective way to convey empathic interest. For example, the patient who was suffering from a sudden paralysis of his whole body angrily told me that his physicians (including me) were “useless” and our treatments “a waste.” I felt threatened and defensive and didn’t know what to say. So I simply repeated his words: “everything we’ve been offering you seems ‘useless,’ and ‘a waste,’” and he felt heard. This led him to tell me much more about what he was going through.

Then, as this patient told me the story of his illness, I found myself relaxing and listening more attentively. One benefit of letting the patient tell his story is that this can literally help the clinician relax when things are tense. One way that all of us relax is by immersing ourselves in another’s story (books, television, movies). Beyond diminishing anxiety, attending to patient narratives predictably initiates empathy. Writing narratives using patient’s own words has been shown to help medical students and residents develop empathic listening skills. Rita Charon and others have shown that clinicians who learn to write such narratives show more empathy for their patients in subsequent clinical work.³⁷

Fifth, body language is as important as words

Recall that observational studies show that patients tell important information to clinicians who are non-verbally attuned. They withhold information from those whose body language does not convey empathy.³⁸ While we cannot directly control much of our body language, physicians can learn to sit down and make eye contact, look up from computer screens and otherwise convey their attentiveness to patients. By taking the steps above to decrease their own anxiety, they will hopefully become more non-verbally attuned as well, as mindfulness improves awareness of others.

Sixth, accepting blame is not a sign of weakness, but can be empowering for both patient and caregiver

Recognizing this will require a cultural shift in which physicians, like other caregivers, learn to be receptive to patients’ feedback, even when this involves blaming the physician. Physicians are not socialized to admit error, and studies show they feel very guilty and inadequate when errors occur.³⁹ During conflicts physicians are likely to become more controlling and less open to patients’ negative feedback. The failure of physicians to acknowledge errors is usually attributed to fears of malpractice, despite research showing that better communication reduces malpractice risk.⁴⁰ Yet even when there is no concern with malpractice physicians rarely acknowledge mistakes, including shortcomings in communication.

Of course there are institutional cultures that influence how clinicians are expected to acknowledge actual medical and technical mistakes. But what I have in mind here are not treatment errors, but interpersonal errors—like being too abrupt or blunt with a patient—which we need not fear acknowledging, yet we rarely apologize. As we discussed in the case of the very sick child, if the pediatric surgeons had apologized for discussing autopsy and organ donation prematurely they might have truly helped the mother express and get support for her grief.

At the right moment, when a patient is very upset, the clinician who allows the patient to express anger towards him or her directly, and can accept blame without becoming defensive, is likely to have great therapeutic impact. For example, during my psychiatry residency we had inadequate resources to offer uninsured patients. Such patients were often appropriately angry about long clinic waits and unavailable clinics. Rather than blaming this lack of resources on “them,” I would directly speak about the limitations of the services that “we” were offering the patients, so that they could directly blame me, and I could then empathize with their appropriate anger.

I’d like to end with another case, one in which we took advantage of the knowledge described above and shifted from sympathetic distress back to engaged curiosity. An oncology team called for an emergency psychiatric consult to evaluate the apparently dangerous nineteen-year-old son of a cancer patient on their ward.⁴¹ His mother was dying and in terrible pain. The son had carried her emaciated body into the hospital after signing her out from two other hospitals against medical advice. He threatened to shoot the nurses who tried to go into his mother’s room to treat her pain, because the pain medication diminished her level of consciousness and he wanted to maintain contact with her. The doctors and nurses were adamant that he was dangerous and should be committed. Yet he had no gun with him, nor was there any indication that he was obtaining a gun, and he had no history of violence.

During a one-hour consultation with the entire medical team, each person became aware of what they found most disturbing about this case. The doctors, nurses and social worker recognized specific fears of their own, involving loss or separation from their children or parents. Their overwhelming feeling of fear about the son’s dangerous behavior shifted to feelings of grief and sadness, and they became calm enough to engage with him with genuine curiosity about what he was going through. He spoke at length with one clinician and told her that he felt terrible about leaving for college across the country the year before when his mother was already sick.

As soon as the medical team members were able to convey their empathy to the young man he stopped making threats. Feeling the team’s support he was able to recognize for the first time that his perspective was distorted by his intense anxiety. This enabled him to phone his father, who, despite his grief, was staying home because he could not deal with his son. The son asked his father to relieve him so that he could take a brief rest (he had not slept for days). When the father came to the hospital, the son began to grieve. When the mother died (comfortably medicated), the father held his son in his arms.

Conclusion: we need a paradigm shift

In closing, we need a paradigm shift in research on clinical empathy. We need to study empathic communication rather than just what goes on inside the minds of physicians. Neuroscience and psychology researchers are just beginning to recognize how their research methodologies (artificially narrow stimuli) segregate cognitive and affective processes that actually work together in life. Recent work shows that the affective expressivity of a target shapes whether an empathic listener uses affective or cognitive processes. By designing studies of clinical empathy that focus on such interpersonal factors we open exciting new possibilities for empowering patients as well as physicians to improve empathic communication.

Finally, we need to question how medical culture, and the institutional and economic forces behind it, impede empathic healthcare. One concrete complaint that I often hear is that physicians do not have time for empathic curiosity. I take this complaint seriously, as I think that the current time and effort demands of practice can be dehumanizing for physicians, and therefore for patients as well. Yet the question of how much more time empathic care takes is still an open one. Many physicians were worried that shifting to a more patient-centered approach by inviting the patient to talk in an open ended way at the beginning of a medical interview would take too much time. Recent research shows that letting patients speak without interrupting usually takes no more than 90 s and improves patient satisfaction and adherence to treatment.⁴²

Of course it is one thing to let a patient speak at the beginning of an interview and another to engage in empathic communication about a patient’s personal needs and concerns. With so little time (20 min primary care visits) and physicians now facing computer screens rather than patients, empathic communication is seriously at risk, and physicians as well as patients are increasingly frustrated.⁴³

Still, I believe that we should resist the tendency to rationalize the current time constraints by presuming that only psychiatrists and social workers need time for genuine empathic communication. This would disserve patients in primary care, including the large number who come in with serious unmet mental health needs that endanger them (15–22 % have major depression).⁴⁴ Recent research shows that even brief empathy interventions during short primary care visits can make a significant difference, reducing the incidence of self-harm and suicide in depressed patients.⁴⁵ This does not mean that such patients did not need to see psychiatrists, but rather that empathic primary care physicians play a crucial role in detecting serious mental health problems and making appropriate referrals.

On the negative side, a lack of skillful emotional communication can set off vicious cycles. Physicians who do not know how to manage patient’s emotional communications are at risk of pathologizing, ignoring, transferring or discharging patients whom they find difficult. This leads to costs in personnel time, legal expenses and patient transfers.

So what are doctors to do? Too often they are not given adequate time to connect with their patients, yet patients sensing this become anxious and dissatisfied, which only makes doctors more anxious and less empathic. This vicious cycle appears to be one of several factors contributing to physicians growing dissatisfaction with their careers—a 2013 survey showed that 59 % would not recommend a medical career.⁴⁶

Still, perhaps shifting our research paradigm to focus more on the interactions between patients and physicians will reveal more about how patients’ responsiveness to physician empathy is conveyed back to physicians, and how this can help physicians experience more meaningfulness and satisfaction in their work.⁴⁷ When physicians and nurses as well as patients see the value they derive from empathic interactions, they may join with other advocates in seeking new models of healthcare with less stringent constraints on the time they have with their patients.

In closing then, the question of whether clinical empathy is primarily cognitive or affective is still open for theoretical debate, but perhaps resolved enough to say the following to current clinicians. It is best when both pathways are engaged, but engaged curiosity and awareness of one’s own emotions are essential. But such questions are no longer the most important ones for improving medical care, for they are still too focused on the mind of the physician and his or her individual traits and capacities, rather than on the mutually shaping interactions of patients and physicians. The most robust findings about clinical empathy come from observational studies in which patients who perceive their physicians as empathic give fuller histories, adhere to treatment, and take steps to treat their diseases. The action here is not in the head of the physician but in the interpersonal field. We need a research agenda that elucidates empathic communication as a sustainable loop between patients and physicians rather than as an additional task for already stressed physicians.